deepcivics 0.1.1

Opensource, LLM powered library to turn public data into civic insight for the Public, Policy makers and Investment professionals.

🏛️ Deep Civics

Open Data + Policy Tools for the People

Deep Civics helps you turn public data into civic insight for the Public, Policy makers and Investment professionals.

It's an open-source Python library and Jupyter interface to let anyone:

• 🗂️ Load and analyze public datasets
• 💬 Ask questions in your language using an LLM
• 📊 Visualize trends, disparities, and impacts
• 🧭 Share findings or use it your applications

🌍 Example Use Cases

• 📉 Understand budget allocation by region
• 🌡️ Compare climate or emissions data over time
• 🏥 Track healthcare data by district
• 🏭 Explore industrial safety or infrastructure trends
• 🏦 Monitor financial and digital inclusion metrics

Quickstart

Let’s use Deep Civics to explore how South Africa has improved digital financial access over time, using World Bank data.

pip install deepcivics

from deepcivics import generate_config_from_url

# Use a real CSV from the World Bank
generate_config_from_url(
    csv_url="https://databankfiles.worldbank.org/public/ddpext_download/ICT/Series/FS.DSR.DIGS.ZS.csv",
    country="South Africa"
)

from deepcivics import CivicLLM
from deepcivics.config import load_config
from deepcivics import ingestion

cfg = load_config()
df = ingestion.fetch_csv(cfg["source"])

llm = CivicLLM(cfg["model"])
context = df.to_csv(index=False)[:cfg["truncate_context"]]
question = "Which countries in Africa have improved digital finance access the most?"
answer = llm.ask(question, context)
print("✅ LLM Answer:", answer)

##📓 Try in Your Browser

##🧠 Architecture

deepcivics/ – Core Python package

notebooks/ – Civic data demos

datasets/ – YAML registry of reusable datasets

docs/ – GitHub Pages documentation

##🤝 Contributing

Y'all are welcome to contribute and provide feedback via PR!

##📜 License

MIT. Use freely, cite responsibly, act respectfully.